Performing Test BLAS_F2C_FLOAT_WORKS - Success Performing Test BLAS_F2C_DOUBLE_WORKS - Failed Performing Test HAS_C11_ATOMICS - Success Performing Test CXX_HAS_AVX2_2 - Success Performing Test CXX_HAS_AVX2_1 - Failed Performing Test CXX_HAS_AVX_2 - Success Performing Test NO_GCC_EBX_FPIC_BUG - Success Performing Test HAVE_GCC_GET_CPUID - Success Could not find hardware support for NEON on this machine. MAGMA LIBRARIES: /root/miniconda/envs/r21d/lib/libmagma.a MAGMA INCLUDE DIRECTORIES: /root/miniconda/envs/r21d/include Checking prototype magma_get_sgeqrf_nb for MAGMA_V2 - True Found CUDA with FP16 support, compiling with CMAKE_MODULE_PATH : /usr/local/pytorch/cmake/Modules /usr/local/pytorch/cmake/public/./Modules_CUDA_fix CMAKE_INSTALL_PREFIX : /root/miniconda/envs/r21d CMAKE_PREFIX_PATH : /root/minoconda/envs/r21d CXX flags : -Wno-deprecated -fvisibility-inlines-hidden -Wnon-virtual-dtor Did anyone have a similar problem and solved it? - ******** Summary ********
OPENCV CMAKE LINUX FULL
That’s the full output I get when compiling it. Using python 2.7, cmake 3.11.4, gcc/g++ 5.4.0, opencv 3.4.4, conda, etc (You can see the rest in the summary below) I’m compiling it with the cuda flag and have a titan x at my disposal. I can get it compiled till the end but when I try to run from caffe2.python import core I get the message saying it doesn’t have GPU support. Well… I’ve tried everything but can’t get it working. I need them compiled together to try to reproduce the environment of a paper that I’m trying to work with. Hey guys, I’m new to pytorch/caffe2 and am having a really hard time getting it compiled from source with the dependencies I need, which are in this case FFMPEG and OPENCV.
0 kommentar(er)
